SSI: The UK Software Sustainability Institute

点击翻译按钮获取中文摘要

This proposal seeks to gain support for the next five years for the UK Software Sustainability Institute (SSI) that will work in partnership with research communities to identify key software that needs to be sustained.Researchers today from many different disciplines rely on software to carry out high quality research. This software must be able to serve the changing needs of the researcher to remain relevant. It must be able to change, adapt, and travel with the researcher on the journey of exploration, innovation and discovery which embodies leading research.Software sustainability is the key to this journey; it is essential that software used in research is managed beyond the lifetime of its original funding cycle, and strengthened, adapted and customised for adoption by researchers within the founding community and outside, in other communities to create a impact which can be transferred to future generations of researchers.The sustainability of important research software requires partnership. Only through close collaboration between software developers and scientists can advances be made both in software development and domain sciences. This is not a radical view. One of the most successful aspects of Agile development methodologies is a focus on ongoing interactions between developers and customers: business people and developers must work together daily throughout the project. The benefits of Agile methods for scientific software development are greater than for business software; and our experiences through our work as OMII-UK underline this emphatically. Through programmes such as ENGAGE and eDIKT2 we have worked in partnership with researchers to improve their software. As each community's requirements are different, the system we have developed is well suited to communities and software of varying maturity: one size does not fit all. We propose a Software Sustainability Institute for the UK which will establish a national focal point - a facility for research software users and developers - based around specialist software engineering skills for driving the continued improvement and impact of research software, using as its foundation the concept of Practical Peer Partnerships: bringing together key research groups within the UK with the skilled team members from our consortium to deliver a series of genuine partnerships focused on the improvement of vital research software through consultative advice, such as test design or help with development tools; collaborative partnerships, e.g. by helping with code refactoring; and long-term engagement, for cultivating the relationships which grow communities and harness the momentum of other activities. This will build on the partners existing connections to many different research communities, and to infrastructure initiatives such as the NGS, EGI, NESSI and PRACE.Our approach will ensure that software used in research is managed beyond the lifetime of its original funding cycle: it will be strengthened, adapted and customised for adoption by researchers, within the founding community and elsewhere, to create an impact which can be built on by future generations of researchers - ultimately to deliver new high quality research.